Home Ribbon Tab¶

The Home ribbon tab in Microscopy Image Browser (MIB) provides tools for handling image datasets, including importing, exporting, saving, and rendering options. This page details each section, with demonstrations and references where available.

Overview¶

The image above shows the Home tab of the ribbon at the top of the MIB interface. This tab organizes essential file-related actions for working with multidimensional microscopy datasets.

Import Image Section¶

Load¶

• Click the Load button to select the current working directory with images. After selection, the image files will be shown in the Directory Contents panel and the current directory in the Status bar will be updated to the selected folder.

• Additionally, a dropdown menu under the button shows the list of recent directories from which images were loaded previously. Selecting a directory from that list updates the current working directory to the selected folder.

Import¶

• MATLAB: Import an image from the main MATLAB workspace into MIB
  Optionally include a dictionary with dataset parameters that were generated by the Export image to command Brief demo
• System Clipboard: Paste an image from the system clipboard
  Uses the IMCLIPBOARD function by Jiro Doke, MathWorks, 2010 Brief demo

• Imaris: Import a dataset from Imaris
  Requires Imaris and ImarisXT and uses IceImarisConnector by Aaron C. Ponti, ETH Zurich
  Demo

• OMERO: connect to an OMERO server and load images.

  Not implemented

  Requires OMERO server files; see System Requirements for installation details.
  Demo

  1. Select a server (do not copy/paste the password):
     
  2. Choose a dataset and range:
     

• URL / Zarr: Open a dataset from a URL address - an ordinary image, or an OME-Zarr container in a public cloud bucket browsed and opened without downloading it (see details)
  The link must include the protocol (e.g., http://) Brief demo

Example datasets¶

Quickly access demo datasets and full DeepMIB projects for image segmentation, grouped by data collection techniques.

DeepMIB projects → Synthetic 2D Large Spots

A complete DeepMIB project with a synthetic dataset for testing semantic segmentation
Includes a trained DeepLabV3-Resnet18 network for detecting large spots on a black background
Load the network via 2D_LargeSpots_2cl_DeepLabV3.mibCfg using:
Ribbon → Tools → Deep learning segmentation → Options tab → Config files → Load

Directory Tree	Segmentation Example

Reference: DOI: 10.5281/zenodo.10203188

DeepMIB projects → Synthetic 2D Small Spots

A complete DeepMIB project with a synthetic dataset for testing semantic segmentation
Includes a trained U-net network for detecting small random spots of two colors on a black background
Load the network via 2D_SmallSpots_3cl_Unet.mibCfg using:
Ribbon → Tools → Deep learning segmentation → Options tab → Config files → Load

Directory Tree	Segmentation Example

Reference: DOI: 10.5281/zenodo.10203764

DeepMIB projects → Synthetic 2.5D Large Spots

A complete DeepMIB project with a synthetic dataset for testing 2.5D depth-to-color semantic segmentation
Includes trained 2.5D DeepLabV3-Resnet18 and U-net networks for segmenting large 3D spots (ignoring 2D spots) using 5-slice subvolumes
Load the networks via:
- Spots_25D_DLv3RN18_Z2C_xy200z5 (DeepLabV3-based)
- Spots_25D_Unet_Z2C_xy200z5 (U-net-based)
Using Ribbon → Tools → Deep learning segmentation → Options tab → Config files → Load or drag-and-drop the config file into DeepMIB

Directory Tree	Segmentation Example

Reference: DOI: 10.5281/zenodo.10212417

DeepMIB projects → Synthetic 2D Patch-wise

A complete DeepMIB project with a synthetic dataset for testing patch-wise segmentation
Includes a trained Resnet18 network for detecting patches of large white spots on a black background
Load the network via 2D_LargeSpots_Patchwise_Resnet18.mibCfg using:
Ribbon → Tools → Deep learning segmentation → Options tab → Config files → Load

Directory Tree	Segmentation Example

Reference: DOI: 10.5281/zenodo.10203861

DeepMIB projects → 2D EM Membranes

A complete DeepMIB project for segmenting membranes from serial-section TEM images
Includes trained U-net and DeepLabV3-Resnet18 networks
Updated project from DeepMIB paper (Figure 1a)
Load via 2D_EM_membranes_Unet.mibCfg or 2D_EM_membranes_DLabRN18.mibCfg using:
Ribbon → Tools → Deep learning segmentation → Options tab → Config files → Load
See readme.txt for details

Directory Tree	Segmentation Example

DeepMIB projects → 2D LM Nuclei

A complete DeepMIB project for segmenting nuclei, boundaries, and touching edges
Includes a trained U-net network
Updated project from DeepMIB paper (Figure 1b)
Load via valid_252px_32patches_50ep.mibCfg using:
Ribbon → Tools → Deep learning segmentation → Options tab → Config files → Load
See readme.txt for details

Directory Tree	Segmentation Example

DeepMIB projects → 3D EM Mitochondria

A complete DeepMIB project for segmenting mitochondria
Includes a trained 3D U-net network
Updated project from DeepMIB paper (Figure 1c)
Load via NoValidation_valid_Aug_128px_256pat.mibCfg using:
Ribbon → Tools → Deep learning segmentation → Options tab → Config files → Load
See readme.txt for details

Directory Tree	Segmentation Example

DeepMIB projects → 3D LM Inner Hair Cells

A complete DeepMIB project for segmenting inner hair cells, nuclei, and synapses
Includes a trained 3D anisotropic U-net network
Updated project from DeepMIB paper (Figure 1d)
Load via InnerEar3D_Hybrid_Same_136x64px_120ep.mibCfg using:
Ribbon → Tools → Deep learning segmentation → Options tab → Config files → Load
See readme.txt for details

Directory Tree	Segmentation Example

Export Image Section¶

Save Image As¶

Save the open dataset to disk in various formats:

Export Image To¶

Export the current dataset to external applications:

• MATLAB: Export includes a dictionary with dataset parameters for re-import into MIB Demo
• Imaris: Export to Imaris (requires Imaris installation)

• Zarr3: Export the dataset as a chunked, pyramidal OME-Zarr v3 store (.zarr3) — readable by MIB as a BigData dataset and by external OME-Zarr–compatible tools

  Export to Zarr3 — dialog settings

  A settings dialog appears after choosing the output path. Defaults are automatically adapted to the open dataset type (WSI vs. 3-D volumetric); see the Smart defaults table at the bottom.

  Setting	Description
  Pyramid levels (0 = auto)	0 = auto: starts at full resolution, adds a level while min(Y, X) / 2 ≥ 256 px, up to 8 levels. Enter 1–12 to force a fixed count.
  Chunk size [Y, X, Z]	Zarr chunk dimensions in pixels. Smaller chunks give finer random-access granularity; larger chunks improve sequential read throughput.
  Shard X-factors [Y, X, Z]	Integer multipliers specifying how many chunks to bundle per axis into one shard file (zarr v3 sharding codec). For example, 4, 4, 1 packs 4 × 4 × 1 = 16 chunks per file, significantly reducing file count for large pyramids. Set any axis to 0 to disable sharding entirely.
  Compression	zstd (fast, excellent ratio — default), gzip (broader tool compatibility), none (fastest write, largest files).
  Downsampling method	Kernel for building coarser pyramid levels: bilinear (fast, smooth — default), nearest (fast, exact integer values), bicubic (slow, sharper edges), median (slow, noise-robust — good for images with outlier pixels, better edge preservation than bilinear), mode (slow, precise — dominant value per block; for categorical label data exported as images).
  Downsampling strategy	How the Z axis is treated when building coarser levels — see below.

  ---

  Downsampling strategy

  • XY only — every pyramid level halves X and Y; Z stays constant. Voxels become progressively more anisotropic at coarser levels. Default for WSI and near-isotropic 3-D data.

  • Anisotropy-preserving — halves XY until the effective XY voxel size would exceed the Z voxel size; from that point onward, both XY and Z are halved together, keeping the voxel aspect ratio close to 1 : 1 at every pyramid level. Recommended for 3-D datasets where the Z step is coarser than the XY pixel size (e.g. confocal, SBEM stacks with vxZ >> vxXY).

  ---

  Smart defaults (computed from the open dataset)

  Dataset type	Chunk [Y, X, Z]	Shard X-factors	Strategy
  WSI (Z ≤ 2 slices or max(Y, X) ≥ 8 000 px)	512 × 512 × 1	4 × 4 × 1	XY only
  3-D, near-isotropic (vxZ < 2 × vxXY)	128 × 128 × 64	4 × 4 × 1	XY only
  3-D, anisotropic (vxZ ≥ 2 × vxXY)	256 × 256 × 16	4 × 4 × 1	Anisotropy-preserving

Make Snapshot¶

Capture a snapshot of the current slice, including all visible objects.
See Make Snapshot for details.

Render Volume¶

Visualize volumes in 3D using three methods:

MIB Rendering (recommended)¶

Hardware-accelerated volume rendering in MIB (since version 2.5, MATLAB R2018b).
Supports downsampling, snapshots, and animations. Updated in MIB 2.84+ (MATLAB R2022b) to render 1-3 color channels with models (MATLAB-only as of 2.84).
See 3D Viewer for details, Demo.

Volume Rendering Engines of MIB

• MIB3 has only the new rendering engine available

MIB 2.84+, R2022b or newer (Volumes 1-3 colors + models)	MIB 2.5, R2018b or newer (1-channel volumes or single material models)
Limitations: - Available only for MIB for MATLAB (as of 2.84) - One volume at a time	Limitations: - One volume at a time (image or model material) - Grayscale only (single channel) - No scale bar

MATLAB Volume Viewer¶

Export to MATLAB's Volume Viewer app
not available in compiled MIB
Demo

3D viewer in Fiji¶

Render via Fiji (requires installation; see System Requirements).
Demo

Additional parameters:

• Reduce the volume down to, max width pixels: Resize dataset (0 for no resizing)
• Smoothing 3D kernel, width: Apply Gaussian blur (0 for no smoothing)
• Invert? [0-no, 1-yes]: Invert for electron microscopy
• Transparency threshold: Set transparency (comma-separated for each channel), adjustable in Fiji's 3D Viewer (Edit → Attributes → Adjust threshold)

Make Movie¶

Save the dataset as a movie file, capturing all visible objects in the image view window.

See Make Movie for details.

I/O Tools Section¶

Batch Processing¶

Design and apply image processing workflows to multiple images automatically.
See Batch Processing for details.

Dataset Chunking¶

Split a large dataset into smaller subvolumes and reassemble them later, or fuse previously cropped datasets. Use the Dataset chunking dropdown to access:

• Chunk dataset: Split the image into smaller chunks for block-based or parallel processing
• Stitch dataset: Reassemble previously chunked subvolumes back into the full image
• Fuse into dataset: Insert previously cropped subvolumes back into the currently open dataset

See Dataset Chunking for details.

Image Shuffling¶

Shuffle files for blind modelling and revert models to original filenames for analysis. Use the Image shuffling dropdown to access:

• Shuffle images: Randomly reorder and rename images to reduce processing bias
• Restore order: Revert images to their original order and filenames

See Image Shuffling for details.

Preferences Section¶

Layout Management¶

Save and restore the arrangement of MIB panels using the Load layout and Save layout buttons.

Load layout options:

• Load local default layout: Restore the layout saved as your personal default (mibDefaultLayout.json located in the MIB preferences folder)
• Load custom layout: Load a layout from a custom file
• Load MIB default layout: Restore the factory default MIB layout

Save layout options:

• Save the current layout as default: Save the current panel arrangement as your personal default
  Stored as mibDefaultLayout.json in the MIB preferences folder (shown at startup as MIB parameters file: ...)
• Save the current layout in a custom file: Save to a custom file for sharing or backup
• Save the current layout as MIB default: Override the factory default layout (affects all users of this MIB installation)

Theme¶

Switch MIB between the light and dark color themes. Follow the system theme uses the theme set in MATLAB. The choice is remembered for the next MIB session.

Preferences¶

Edit MIB preferences, including colors for Selection, Model, and Mask layers, mouse wheel behavior, key settings, and Undo options.
See Preferences for details.

• Make override default settings file: Save your changes to the default settings as the starting settings of new users of this computer or of this MIB installation, see Overriding default settings

MIB saves configuration in a file generated upon closing:

Help¶

Access help and application information via the Help dropdown:

• Open MIB help: Open the MIB documentation website
• Tip of the day: Display a random usage tip
• Support on image.sc: Open the image.sc community forum for MIB questions and support
• Personal support session: Request a one-on-one remote support session with the MIB team
• Check for update: Check for a newer version of MIB and download it if available
• Your personal stats: View your cumulative MIB usage statistics
• Licenses: View licenses for MIB and all included third-party tools
• About MIB: Show MIB version information and credits

Dev Corner Section¶

Developer mode¶

When enabled, widget tooltips display the internal handle name of each ribbon control - useful for identifying widgets when customizing or scripting MIB.

API class reference¶

Opens the MIB API class reference documentation.

Development¶

Calls controllers.MibRibbon.homeDevTest_Callback() — a reserved entry point for internal debugging. Can also be invoked directly from the MATLAB command line:

mib.cRibbon.homeDevTest_Callback();

Back to MIB | User interface | Ribbon